The NVIDIA RTX 4000 Ada Generation is the objective winner, securing victories in nine out of fourteen comparable specifications. Its most significant advantage over the NVIDIA RTX 2070 SUPER is its FP32 performance, which is 195% higher. While the older card maintains a lead in specific architectural counts, the newer model demonstrates a clear superiority in raw computational throughput across the majority of the measured technical categories.

Users prioritizing raw FP32 performance should select the NVIDIA RTX 4000 Ada Generation for its substantial lead in that metric. Conversely, the NVIDIA RTX 2070 SUPER remains the preferred choice for those who specifically require a higher count of Tensor Cores, where it holds a 67% advantage. Your decision should depend on whether your workload demands the superior computational power of the newer architecture or the specific core configuration offered by the older model.